{ "_attn_implementation_autoset": false, "acoustic_tokenizer_config": { "causal": true, "channels": 1, "conv_bias": true, "conv_norm": "none", "corpus_normalize": 0.0, "decoder_depths": null, "decoder_n_filters": 32, "decoder_ratios": [ 8, 5, 5, 4, 2, 2 ], "disable_last_norm": true, "dtype": "bfloat16", "encoder_depths": "3-3-3-3-3-3-8", "encoder_n_filters": 32, "encoder_ratios": [ 8, 5, 5, 4, 2, 2 ], "fix_std": 0.5, "layer_scale_init_value": 1e-06, "layernorm": "RMSNorm", "layernorm_elementwise_affine": true, "layernorm_eps": 1e-05, "mixer_layer": "depthwise_conv", "model_type": "vibevoice_acoustic_tokenizer", "pad_mode": "constant", "std_dist_type": "gaussian", "vae_dim": 64, "weight_init_value": 0.01 }, "acoustic_vae_dim": 64, "architectures": [ "VibeVoiceASRForConditionalGeneration" ], "decoder_config": { "attention_dropout": 0.0, "bos_token_id": null, "dtype": "bfloat16", "eos_token_id": null, "hidden_act": "silu", "hidden_size": 3584, "initializer_range": 0.02, "intermediate_size": 18944, "layer_types": [ "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention" ], "max_position_embeddings": 131072, "max_window_layers": 28, "model_type": "qwen2", "num_attention_heads": 28, "num_hidden_layers": 28, "num_key_value_heads": 4, "pad_token_id": null, "rms_norm_eps": 1e-06, "rope_parameters": { "rope_theta": 1000000.0, "rope_type": "default" }, "sliding_window": null, "tie_word_embeddings": false, "use_cache": true, "use_mrope": false, "use_sliding_window": false, "vocab_size": 152064 }, "diffusion_head_config": { "ddpm_batch_mul": 4, "ddpm_beta_schedule": "cosine", "ddpm_num_inference_steps": 20, "ddpm_num_steps": 1000, "diffusion_type": "ddpm", "head_ffn_ratio": 3.0, "head_layers": 4, "hidden_size": 3584, "latent_size": 64, "model_type": "vibepod_diffusion_head", "prediction_type": "v_prediction", "rms_norm_eps": 1e-05, "speech_vae_dim": 64 }, "dtype": "bfloat16", "model_type": "vibevoice", "quantization_config": { "config_groups": { "group_0": { "format": "nvfp4-pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": "local", "group_size": 16, "num_bits": 4, "observer": "static_minmax", "observer_kwargs": {}, "scale_dtype": "torch.float8_e4m3fn", "strategy": "tensor_group", "symmetric": true, "type": "float", "zp_dtype": null }, "output_activations": null, "targets": [ "re:.*mlp.gate_proj$", "re:.*mlp.up_proj$", "re:.*mlp.down_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": 16, "num_bits": 4, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": "torch.float8_e4m3fn", "strategy": "tensor_group", "symmetric": true, "type": "float", "zp_dtype": null } } }, "format": "nvfp4-pack-quantized", "global_compression_ratio": null, "ignore": [ "re:.*self_attn.*", "model.language_model.layers.0.self_attn.q_proj", "model.language_model.layers.0.self_attn.k_proj", "model.language_model.layers.0.self_attn.v_proj", "model.language_model.layers.0.self_attn.o_proj", "model.language_model.layers.1.self_attn.q_proj", "model.language_model.layers.1.self_attn.k_proj", "model.language_model.layers.1.self_attn.v_proj", "model.language_model.layers.1.self_attn.o_proj", "model.language_model.layers.2.self_attn.q_proj", "model.language_model.layers.2.self_attn.k_proj", "model.language_model.layers.2.self_attn.v_proj", "model.language_model.layers.2.self_attn.o_proj", "model.language_model.layers.3.self_attn.q_proj", "model.language_model.layers.3.self_attn.k_proj", "model.language_model.layers.3.self_attn.v_proj", "model.language_model.layers.3.self_attn.o_proj", "model.language_model.layers.4.self_attn.q_proj", "model.language_model.layers.4.self_attn.k_proj", "model.language_model.layers.4.self_attn.v_proj", "model.language_model.layers.4.self_attn.o_proj", "model.language_model.layers.5.self_attn.q_proj", "model.language_model.layers.5.self_attn.k_proj", "model.language_model.layers.5.self_attn.v_proj", "model.language_model.layers.5.self_attn.o_proj", "model.language_model.layers.6.self_attn.q_proj", "model.language_model.layers.6.self_attn.k_proj", "model.language_model.layers.6.self_attn.v_proj", "model.language_model.layers.6.self_attn.o_proj", "model.language_model.layers.7.self_attn.q_proj", "model.language_model.layers.7.self_attn.k_proj", "model.language_model.layers.7.self_attn.v_proj", "model.language_model.layers.7.self_attn.o_proj", "model.language_model.layers.8.self_attn.q_proj", "model.language_model.layers.8.self_attn.k_proj", "model.language_model.layers.8.self_attn.v_proj", "model.language_model.layers.8.self_attn.o_proj", "model.language_model.layers.9.self_attn.q_proj", "model.language_model.layers.9.self_attn.k_proj", "model.language_model.layers.9.self_attn.v_proj", "model.language_model.layers.9.self_attn.o_proj", "model.language_model.layers.10.self_attn.q_proj", "model.language_model.layers.10.self_attn.k_proj", "model.language_model.layers.10.self_attn.v_proj", "model.language_model.layers.10.self_attn.o_proj", "model.language_model.layers.11.self_attn.q_proj", "model.language_model.layers.11.self_attn.k_proj", "model.language_model.layers.11.self_attn.v_proj", "model.language_model.layers.11.self_attn.o_proj", "model.language_model.layers.12.self_attn.q_proj", "model.language_model.layers.12.self_attn.k_proj", "model.language_model.layers.12.self_attn.v_proj", "model.language_model.layers.12.self_attn.o_proj", "model.language_model.layers.13.self_attn.q_proj", "model.language_model.layers.13.self_attn.k_proj", "model.language_model.layers.13.self_attn.v_proj", "model.language_model.layers.13.self_attn.o_proj", "model.language_model.layers.14.self_attn.q_proj", "model.language_model.layers.14.self_attn.k_proj", "model.language_model.layers.14.self_attn.v_proj", "model.language_model.layers.14.self_attn.o_proj", "model.language_model.layers.15.self_attn.q_proj", "model.language_model.layers.15.self_attn.k_proj", "model.language_model.layers.15.self_attn.v_proj", "model.language_model.layers.15.self_attn.o_proj", "model.language_model.layers.16.self_attn.q_proj", "model.language_model.layers.16.self_attn.k_proj", "model.language_model.layers.16.self_attn.v_proj", "model.language_model.layers.16.self_attn.o_proj", "model.language_model.layers.17.self_attn.q_proj", "model.language_model.layers.17.self_attn.k_proj", "model.language_model.layers.17.self_attn.v_proj", "model.language_model.layers.17.self_attn.o_proj", "model.language_model.layers.18.self_attn.q_proj", "model.language_model.layers.18.self_attn.k_proj", "model.language_model.layers.18.self_attn.v_proj", "model.language_model.layers.18.self_attn.o_proj", "model.language_model.layers.19.self_attn.q_proj", "model.language_model.layers.19.self_attn.k_proj", "model.language_model.layers.19.self_attn.v_proj", "model.language_model.layers.19.self_attn.o_proj", "model.language_model.layers.20.self_attn.q_proj", "model.language_model.layers.20.self_attn.k_proj", "model.language_model.layers.20.self_attn.v_proj", "model.language_model.layers.20.self_attn.o_proj", "model.language_model.layers.21.self_attn.q_proj", "model.language_model.layers.21.self_attn.k_proj", "model.language_model.layers.21.self_attn.v_proj", "model.language_model.layers.21.self_attn.o_proj", "model.language_model.layers.22.self_attn.q_proj", "model.language_model.layers.22.self_attn.k_proj", "model.language_model.layers.22.self_attn.v_proj", "model.language_model.layers.22.self_attn.o_proj", "model.language_model.layers.23.self_attn.q_proj", "model.language_model.layers.23.self_attn.k_proj", "model.language_model.layers.23.self_attn.v_proj", "model.language_model.layers.23.self_attn.o_proj", "model.language_model.layers.24.self_attn.q_proj", "model.language_model.layers.24.self_attn.k_proj", "model.language_model.layers.24.self_attn.v_proj", "model.language_model.layers.24.self_attn.o_proj", "model.language_model.layers.25.self_attn.q_proj", "model.language_model.layers.25.self_attn.k_proj", "model.language_model.layers.25.self_attn.v_proj", "model.language_model.layers.25.self_attn.o_proj", "model.language_model.layers.26.self_attn.q_proj", "model.language_model.layers.26.self_attn.k_proj", "model.language_model.layers.26.self_attn.v_proj", "model.language_model.layers.26.self_attn.o_proj", "model.language_model.layers.27.self_attn.q_proj", "model.language_model.layers.27.self_attn.k_proj", "model.language_model.layers.27.self_attn.v_proj", "model.language_model.layers.27.self_attn.o_proj", "model.acoustic_tokenizer.encoder.stages.0.0.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.0.0.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.0.1.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.0.1.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.0.2.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.0.2.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.1.0.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.1.0.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.1.1.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.1.1.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.1.2.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.1.2.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.2.0.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.2.0.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.2.1.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.2.1.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.2.2.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.2.2.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.3.0.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.3.0.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.3.1.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.3.1.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.3.2.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.3.2.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.4.0.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.4.0.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.4.1.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.4.1.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.4.2.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.4.2.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.5.0.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.5.0.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.5.1.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.5.1.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.5.2.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.5.2.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.6.0.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.6.0.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.6.1.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.6.1.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.6.2.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.6.2.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.6.3.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.6.3.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.6.4.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.6.4.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.6.5.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.6.5.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.6.6.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.6.6.ffn.linear2", "model.acoustic_tokenizer.encoder.stages.6.7.ffn.linear1", "model.acoustic_tokenizer.encoder.stages.6.7.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.0.0.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.0.0.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.0.1.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.0.1.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.0.2.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.0.2.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.0.3.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.0.3.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.0.4.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.0.4.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.0.5.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.0.5.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.0.6.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.0.6.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.0.7.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.0.7.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.1.0.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.1.0.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.1.1.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.1.1.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.1.2.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.1.2.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.2.0.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.2.0.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.2.1.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.2.1.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.2.2.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.2.2.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.3.0.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.3.0.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.3.1.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.3.1.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.3.2.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.3.2.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.4.0.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.4.0.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.4.1.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.4.1.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.4.2.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.4.2.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.5.0.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.5.0.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.5.1.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.5.1.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.5.2.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.5.2.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.6.0.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.6.0.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.6.1.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.6.1.ffn.linear2", "model.acoustic_tokenizer.decoder.stages.6.2.ffn.linear1", "model.acoustic_tokenizer.decoder.stages.6.2.ffn.linear2", "model.semantic_tokenizer.encoder.stages.0.0.ffn.linear1", "model.semantic_tokenizer.encoder.stages.0.0.ffn.linear2", "model.semantic_tokenizer.encoder.stages.0.1.ffn.linear1", "model.semantic_tokenizer.encoder.stages.0.1.ffn.linear2", "model.semantic_tokenizer.encoder.stages.0.2.ffn.linear1", "model.semantic_tokenizer.encoder.stages.0.2.ffn.linear2", "model.semantic_tokenizer.encoder.stages.1.0.ffn.linear1", "model.semantic_tokenizer.encoder.stages.1.0.ffn.linear2", "model.semantic_tokenizer.encoder.stages.1.1.ffn.linear1", "model.semantic_tokenizer.encoder.stages.1.1.ffn.linear2", "model.semantic_tokenizer.encoder.stages.1.2.ffn.linear1", "model.semantic_tokenizer.encoder.stages.1.2.ffn.linear2", "model.semantic_tokenizer.encoder.stages.2.0.ffn.linear1", "model.semantic_tokenizer.encoder.stages.2.0.ffn.linear2", "model.semantic_tokenizer.encoder.stages.2.1.ffn.linear1", "model.semantic_tokenizer.encoder.stages.2.1.ffn.linear2", "model.semantic_tokenizer.encoder.stages.2.2.ffn.linear1", "model.semantic_tokenizer.encoder.stages.2.2.ffn.linear2", "model.semantic_tokenizer.encoder.stages.3.0.ffn.linear1", "model.semantic_tokenizer.encoder.stages.3.0.ffn.linear2", "model.semantic_tokenizer.encoder.stages.3.1.ffn.linear1", "model.semantic_tokenizer.encoder.stages.3.1.ffn.linear2", "model.semantic_tokenizer.encoder.stages.3.2.ffn.linear1", "model.semantic_tokenizer.encoder.stages.3.2.ffn.linear2", "model.semantic_tokenizer.encoder.stages.4.0.ffn.linear1", "model.semantic_tokenizer.encoder.stages.4.0.ffn.linear2", "model.semantic_tokenizer.encoder.stages.4.1.ffn.linear1", "model.semantic_tokenizer.encoder.stages.4.1.ffn.linear2", "model.semantic_tokenizer.encoder.stages.4.2.ffn.linear1", "model.semantic_tokenizer.encoder.stages.4.2.ffn.linear2", "model.semantic_tokenizer.encoder.stages.5.0.ffn.linear1", "model.semantic_tokenizer.encoder.stages.5.0.ffn.linear2", "model.semantic_tokenizer.encoder.stages.5.1.ffn.linear1", "model.semantic_tokenizer.encoder.stages.5.1.ffn.linear2", "model.semantic_tokenizer.encoder.stages.5.2.ffn.linear1", "model.semantic_tokenizer.encoder.stages.5.2.ffn.linear2", "model.semantic_tokenizer.encoder.stages.6.0.ffn.linear1", "model.semantic_tokenizer.encoder.stages.6.0.ffn.linear2", "model.semantic_tokenizer.encoder.stages.6.1.ffn.linear1", "model.semantic_tokenizer.encoder.stages.6.1.ffn.linear2", "model.semantic_tokenizer.encoder.stages.6.2.ffn.linear1", "model.semantic_tokenizer.encoder.stages.6.2.ffn.linear2", "model.semantic_tokenizer.encoder.stages.6.3.ffn.linear1", "model.semantic_tokenizer.encoder.stages.6.3.ffn.linear2", "model.semantic_tokenizer.encoder.stages.6.4.ffn.linear1", "model.semantic_tokenizer.encoder.stages.6.4.ffn.linear2", "model.semantic_tokenizer.encoder.stages.6.5.ffn.linear1", "model.semantic_tokenizer.encoder.stages.6.5.ffn.linear2", "model.semantic_tokenizer.encoder.stages.6.6.ffn.linear1", "model.semantic_tokenizer.encoder.stages.6.6.ffn.linear2", "model.semantic_tokenizer.encoder.stages.6.7.ffn.linear1", "model.semantic_tokenizer.encoder.stages.6.7.ffn.linear2", "model.acoustic_connector.fc1", "model.acoustic_connector.fc2", "model.semantic_connector.fc1", "model.semantic_connector.fc2", "lm_head" ], "kv_cache_scheme": null, "quant_method": "compressed-tensors", "quantization_status": "compressed", "sparsity_config": {}, "transform_config": {}, "version": "0.15.1.dev14+g01a1c9a" }, "semantic_tokenizer_config": { "causal": true, "channels": 1, "conv_bias": true, "conv_norm": "none", "corpus_normalize": 0.0, "disable_last_norm": true, "dtype": "bfloat16", "encoder_depths": "3-3-3-3-3-3-8", "encoder_n_filters": 32, "encoder_ratios": [ 8, 5, 5, 4, 2, 2 ], "fix_std": 0, "layer_scale_init_value": 1e-06, "layernorm": "RMSNorm", "layernorm_elementwise_affine": true, "layernorm_eps": 1e-05, "mixer_layer": "depthwise_conv", "model_type": "vibevoice_semantic_tokenizer", "pad_mode": "constant", "std_dist_type": "none", "vae_dim": 128, "weight_init_value": 0.01 }, "semantic_vae_dim": 128, "transformers_version": "5.2.0" }